A feasibility study on importance of using Big data Analysis in implementing E-Government services

,

coefficient and T-statistic have been used to test the hypotheses, which according to the obtained results, the confirmed hypotheses should be higher than 1.96.The statistical population of this study is the managers of different levels of Islamic Azad University in Khorasan Razavi province who have been selected by simple stratified random sampling method.Considering that the statistical population of this research is 235 people, Cochran's formula has been used to determine the sample size.

2-Related works
In Keoduangsine et al.'s research [2], a new model was reviewed in order to identify the most influential factors based on Ruff theory analysis and it was found that infrastructure investments are considered as the most effective factor in the development of e-government readiness.In [3] a study is presented based on qualitative analysis approaches using content analysis that created a coherent structure in decision-making level of education in school systems.In this study, three structural changes including training, evaluation and accreditation are considered as the most important educational platforms for data production in educational institutions that can change the decision-making methods in educational institutions.In [4] an approach which was conducted in the period of one year 2017-2018, 400 managers of active companies in the field of private service sector in China participated.In this research, partial least squares analysis (PLS) method was used to fit the model.The results showed that increasing the level of dynamism in recognizing technological management styles and big data analysis in line with the dynamic capabilities of companies can help to develop and effectiveness of decision-making level and cause the level of quality services to be developed.Shonhe et al. [5] is focused on information technology participation.The analysis method in their study was basic data analysis based on three stages of open coding, selective coding and axial coding.The results showed that the use of social media, use of update networks and websites with useful information content for consumers can increase the level of interaction of the audience.Torrecilla et al. [6] was attended by 104 IT managers and R&D managers in the automotive industry.In this study, which was conducted over a period of one year, the method of experimental analysis using the experimental and control groups was used.One group was examined from the perspective of big data analysis theory only and one group was examined after passing a 6-month course.The results clearly showed the difference in the effectiveness of learning between the experimental and control groups, so that the group that had received practical training used more knowledge-based approaches in analyzing the problems of the organization.Patel et al.'s method [1] was based on simulation analysis, examined more than 30 scenarios at specific time intervals based on characteristics such as the type of governance, how to manage resources, the effectiveness of optimal allocation, and so on.The results showed that optimizing the data storage process in the form of a comprehensive database while increasing the speed of information flow increases the level of transparency.It was also found that big data analysis is very important to gain insight and predict the future of government governance, but extensive infrastructure investments need to be made in this regard.In [7] a standard questionnaire was used to collect research data and participants Research 117 small company managers had an activity life of less than 5 years.The results showed that there is a direct relationship between technical training, knowledge training and human resource skills training with the effectiveness of IT implementation, but it turned out that investing in IT infrastructure in small companies, at least for the short term, does not affect the provision of effective technological services in these companies.

3-Research Method
The validity of the questionnaires of this research was standardized using the opinions of experts and specialists.To evaluate the reliability of the research, Cronbach's alpha method was used according to the following equation: where x represents the reliability coefficient of the whole test, k represents the number of test questions, x 2 is the variance of the scores of question j, and x 3 is the variance of the scores of the total questions.If this coefficient is greater than 0.7, the test has acceptable reliability.The Cronbach's alpha value is as shown in Table .1. Therefore, considering that the alpha value for all variables of the questionnaire is greater than 0.7, the reliability of the questionnaire is confirmed.According to the hypotheses presented and the statistical method used, the research findings are presented in the form of descriptive statistics tables and inferential statistics.For this purpose, first the descriptive findings are expressed, then in order to test the hypotheses, the partial least squares analysis (PLS) method is used, which is based on model fitting based on the validity and reliability.Demographic statistics are used to explain the data collected in the research.The frequency distribution of the respondents based on their education, age and managerial level is shown in  Examining the average of the calculated variables, it was found that the highest average is related to the big data diversity variable (4.39), which indicates that the existence of different levels of data in the form of technological software can help improve the level of analytical efficiency and cause Provide more reliable and reliable information for users.It was also found that the core of government services (4/29) is considered as one of the consequences of implementing e-government services, which can reflect the nature and philosophy of services that can be provided and the need for it in the community.The highest rate of standard deviation is related to the speed of big data, which with a deviation of about (0.95) indicates that the participants in this study had different views on the speed of big data in the analysis and service delivery, and this issue It has increased the level of standard deviation.One of the tests used to determine the normality of data distribution is the Kolmogorov-Smirnov test, which is a non-parametric test.In this test, if the significant level of Z statistic value is greater than 5% (Sig <0.05), the statistical null hypothesis that the distribution of the variable under study is normal is accepted with 95% confidence, but otherwise the hypothesis is confirmed that States that the data is not normally distributed.Examination of the results of Kolmogorov-Smirnov test showed that the significance level of Z statistic of all research variables is significant at the 5% error level (Sig <0.05), hence the percentage assumption that the data is normal is rejected and the acceptance hypothesis is rejected.According to the results of Table .6 and considering that the structural equation modeling method with partial least squares (Smart-Pls) approach has been used to analyze the research data, and in this method, the absence of normal data distribution does not create a limit.The model is modeled in the form of validity and reliability of dimensions.It is noteworthy that the software that uses structural equation modeling based on this statistical method is compatible with conditions such as the alignment of independent variables, the normality of the data and the small size of the sample.Table .6 shows the correlation matrix of the main components of the research model.Based on this matrix, since the correlation between the model variables is less than 0.9, so the non-overlap between the research model variables in the form of differential validity is also confirmed.
Table.1 shows the Cronbach's alpha coefficient of research variables.

Table . 1 Cronbach's alpha coefficient of research variables
In this study, stratified random sampling method was used.In this method, based on the number of each class and the percentage of the specified sample volume, the sample is determined.According to the statistics and information announced by the Ministry of Science of Iran, there are 12 branches of the Islamic Azad University in the province of Khorasan Razavi, whose total number of managers is equal to 235 people at different levels.Based on Cochran's formula, the sample size required for this study was 146 samples.